cells
Molecular Plant-Microbe Interactions (Submitted)
Emmanuel Jourdan1*, Guillaume Henry2*, Francéline Duby3, Jacques Dommes3, Philippe Thonart1,2 and Marc Ongena2
* Authors contributed equally to this article
1 Centre Wallon de Biologie Industrielle, Université of Liège, Belgium
2 Unité de Bioindustries, Faculty of Agricultural Sciences of Gembloux, Belgium
3 Laboratoire de Biologie Moléculaire et Biotechnologie Végétales, Université of Liège, Belgium
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
149 Résumé
Dans les chapitres précédents, nous avons démontré que la surfactine et dans une moindre mesure la fengycine, toutes deux produites par Bacillus subtilis S499 étaient impliquées dans l’établissement de l’ISR contre Botrytis cinerea. De plus, les analyses des réponses induites dans la plante avaient montré une stimulation de la lipoxygénase dans les feuilles des plants de tomate traités. Toujours dans l’optique de mieux comprendre les interactions entre les cellules des plantes et les éliciteurs bactériens, nous avons logiquement entrepris lors des travaux présentés dans ce chapitre, d‘entamer une première caractérisation des événements précoces déclenchés par leur reconnaissance et de tenter d’identifier les groupements structuraux des LPs nécessaires à cette reconnaissance. Pour se faire, nous avons utilisé le modèle des cellules de tabac en culture in vitro.
L’addition de surfactine dans les suspensions de cellules déclenche une rapide alcalinisation du milieu extracellulaire suivie par une production d’espèces oxydantes, sans toutefois provoquer de mort cellulaire. Nous avons aussi mis en évidence une concentration seuil en surfactine, de 4 µM, pour visualiser une réponse optimale des cellules de tabac.
L’ajout d’EDTA ou de LaCl3, respectivement un chélateur du calcium et un inhibiteur de canaux calcique, avant le traitement par la surfactine inhibe fortement l’alcalinisation du pH, suggérant l’intervention de l’ion Ca2+ dans la mise en place d’une réponse par les cellules.
D’autres part, la surfactine est aussi en mesure de stimuler l’activité d’enzymes comme la LOX ou la PAL quelques heures après son ajout dans les cultures pour atteindre des valeurs maximales entre 9 et 12h après le traitement. Des analyses en biologie moléculaire ont également révélé une accumulation claire d’ARNm de PAL au même moment que les pics d’activité, illustrant une transcription effective des gènes codant pour l’enzyme causée par la perception de la surfactine. Cette voie de défense a été par la suite approfondie par la mise en évidence de modification dans le profil des phénylpropanoïdes en HPLC et HPLC-MS. Une apparition d’acide salicylique et l’augmentation des taux d’acide cinnamique et hydroxy-benzoïque, ainsi qu’une diminution de ceux des acides ferulique et coumarique ont été observés, et suggèrent alors une réorientation de la voie de phénylpropanoïdes.
Des réponses différentes ont cependant été obtenu avec la fengycine et sont présentées dans les résultats additionnels. Lors de son addition dans les cultures à des concentrations similaires à la surfactine, aucune variation du pH ni de production d’espèces oxydantes n’a été observée dans les délais analysés. Mais une réponse positive a aussi été mise en évidence au niveau enzymatique, puisque son ajout est accompagné par une augmentation de l’activité PAL ainsi qu’une accumulation des transcrits correspondant à cette enzyme.
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
150
Enfin, le mode de reconnaissance et les sous-structures de la surfactine impliquées dans son activité élicitrice ont été investigués grâce à des homologues possédants des longueurs variables de chaînes lipidiques et à des lipopeptides structurellement modifiés par linéarisation ou méthylation au niveau de la partie peptidique. Les surfactines méthylées et linéaires sont moins actives, suggérant respectivement un rôle des charges présentes sur la chaîne peptidique et une importance de la structure cyclique de la molécule. De plus, une activité supérieure a été observée pour les homologues ayant des chaînes lipidiques constituées de 14 ou 15 carbones, en comparaison à ceux possédant des chaînes plus courtes. L’ensemble de ses données suggère que l’action élicitrice de la surfactine sur les cellules de tabac puisse être dépendante de ses propriétés amphiphiles et est discutée dans ce chapitre.
Abstract
Recognition by plant cells of some molecular patterns harboured by microbial pathogens can trigger the activation of defence pathways and eventually lead to a systemic increase of resistance to subsequent attack. In the context of systemic resistance induced by non-pathogenic rhizobacteria, we have recently demonstrated that surfactins and fengycins lipopeptides from Bacillus subtilis S499, were involved in elicitation of the phenomenon in bean and tomato. Here, we further investigated molecular events underlying this ISR-related interaction between such lipopeptides and plants. Addition of surfactin in the micromolar range to tobacco cell suspensions clearly induced some defence-related early events such as Ca2+ -dependent pH alkalinization and reactive oxygen species production but only limited cell death. Surfactin also stimulated defence enzymes such as phenylalanine amonia lyase and lipoxygenase, and modified the pattern of phenolics produced by the elicited cells. Reduced activity of some homologues also indicates that surfactin perception is dictated by structural clues both in the acyl moiety and in the cyclic peptide part. The present study sheds new light not only on defence-related events induced following recognition of amphiphilic lipopeptides from Bacillus but also more globally on the way elicitors from beneficial bacteria can be perceived by host plant cells.
Introduction
During evolution, plants have adapted themselves to environmental conditions. Beyond abiotic stress such as to low or to high temperatures, dryness, high salinity or wounding, plants have to
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
151 protect themselves against diseases caused by a wide range of microorganisms. To develop diseases, these pathogens have to penetrate the plant tissue, either by penetrating leave or root epidermis, or by entering through natural opening such as stomata (Chisholm et al., 2006). To prevent these invasions, plants have evolved several defence strategies. Preformed physical barriers such as cell wall, or constitutively produced antimicrobial compounds can slow down or inhibit pathogen colonization. In a more active way, plants can also recognize some pathogen-associated molecular patterns (PAMPs). These general elicitors of either oligosaccharide, peptide, or lipid nature are non-specific compounds constitutively produced by the pathogen (Gómez-Gómez, 2004). They are involved in indispensable functions for the microorganism such as mobility (flagellin), enzymatic activities (Pep13) or cell surface composition (lipopolysaccharides (LPS), oligosaccharides) (Montesano et al., 2003). PAMPs are recognized by associated- cell surface receptors (Jones and Dangl, 2006). In another way, plant can also recognize cellular changes caused by Avr proteins, some specific pathogen effectors which block signal transduction pathway triggered by general elicitors (Chisholm et al., 2006; Van Loon et al., 2006). Recognition of Avr proteins by plant resistance proteins (R proteins) activates a variety of defence response including the hypersensitive response (HR) (Heath, 2000). HR is characterized by a rapid production of oxidative species, a programmed cell death around the infection site and the synthesis of pathogenesis related proteins (PR proteins), involved in defence against pathogen (Greenberg and Yao, 2004). HR will restrict pathogen growth and confer a kind of resistance in local tissues (Kombrink and Schmelzer, 2001). By contrast to these local responses, plants can develop a systemic form of resistance that extends to all organs following localized interaction. In this context, the systemic acquired resistance (SAR) is the best characterized phenomenon. SAR is activated after a first infection by an incompatible necrotizing pathogen and renders the host plant more resistant to a subsequent attack by a range of virulent pathogens on the same or another organ (Durrant and Dong, 2004; Sticher, 1997). A systemic immunization decreasing disease impact can be triggered by specific strains of plant growth promoting rhizobacteria. This phenomenon has been termed PGPR-induced systemic resistance (ISR) and is also effective against a broad range of fungal, bacterial, and viral diseases, as well as against some insect and nematode pests (Ongena and Thonart, 2006; Van Loon et al., 1998). Several beneficial rhizobacteria have been isolated for their ISR-inducing activities and many of them belong to the Pseudomonas and Bacillus genera (Bakker et al., 2007; Kloepper et al., 2004).
ISR can be globally viewed as a three-step process: bacterial elicitor perception, systemic signal transduction and defence gene expression leading to enhanced responsive capacity of the host plant. ISR-associated signal transduction and defence mechanisms are being well documented even if comparatively less well understood than in the case of SAR (Van Loon
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
152
and Bakker, 2005; Pieterse et al., 2001). By contrast, very little is known about the molecular events governing the early interaction between the bacteria and the plant cell. In some cases, molecules responsible for the ISR-eliciting activity of these PGPR strains have been characterized. These elicitors may be i) cell surface components such as outer-membrane lipopolysaccharides (Reitz et al., 2002; Coventry and Dubery, 2001; Duijff et al., 1997) and flagellin from Pseudomonas species (Meziane et al., 2005), ii) iron regulated metabolites such as pyoverdins, pyochelin, salicylic acid or a N-alkylated benzylamine derivative (Ran et al., 2005; Audenaert et al., 2002; De Meyer et al., 1999; Leeman et al., 1996), iii) the antibiotic compounds DAPG, pyocyanin or massetolide A (De Vleesschauwer et al., 2006; Raaijmakers et al., 2006; Iavicoli et al., 2003; Siddiqui and Shaukat, 2003). In addition, volatiles (2,3-butanediol) and quorum sensing signals (N-homoserinelactone) were recently described as new active compounds capable to induce ISR (Schuhegger et al., 2006; Ryu et al., 2004).
Recent advances have highlighted the important role played by cyclic lipopeptides (LPs) of the iturin, fengycin and surfactin families in plant disease control by various Bacillus strains.
Bacillus lipopeptides (LPs) were at first mostly studied for their antagonistic activity against a wide range of potential phytopathogens including viruses, bacteria, fungi and oomycetes (Haas and Défago, 2005; Leclère et al., 2005; Ongena et al., 2005a; Zahir et al., 2004). It is now clear that such compounds may act not only as antagonists but also by facilitating root colonization (Bais et al., 2004; Hofemeister et al., 2004) and by reinforcing the host resistance potential (Ongena et al., 2005b). In a recent work, we have indeed demonstrated that surfactins and fengycins but not iturins produced by Bacillus subtilis S499 are involved in ISR elicitation on bean and tomato (Ongena et al., 2007).
Fengycins are cyclic lipodecapeptides with a ß-hydroxy fatty acid chain saturated or not with a length of C14 to C18 (Schneider et al., 1999; Vanittanakom et al., 1986). The surfactin family encompasses structural variants isolated from various Bacillus species but all members are heptapeptides interlinked with a ß-hydroxy fatty acid to form a cyclic lactone ring structure (Peypoux et al., 1999). They are among the most powerful biosurfactant known with exceptional emulsifying and foaming properties. Because of their amphiphilic nature, surfactins can also readily associate and tightly anchor into lipid layers (Heerklotz and Seelig, 2007; Deleu et al., 2003; Sheppard et al., 1991).
By contrast to the numerous investigations conducted with some PAMPs used as models for the study of early defence-related events (Garcia-Brugger et al., 2006; Zhao et al., 2005;
Gómez-Gómez, 2004), very few informations are available about the perception mechanisms of ISR-specific elicitors by plant cells. The general objective of the work presented here was to provide a first picture of the metabolic changes that can be induced in plant cells upon recognition of elicitors from non-pathogenic bacteria. More specifically, we have investigated
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
153 the responses to surfactins in cultivated tobacco cells and have identified some early events as well as later biochemical changes that could ultimately lead to an enhanced state of resistance.
Results
Surfactin as main inducer of the alkalinization response of tobacco cells
As a member of the Bacillus genus, the rhizobacterium B. subtilis has the potential to produce a vast array of biologically active molecules among which structurally diversified anti-microbial compounds (Stein, 2005). The first objective of this work was thus to evaluate whether other molecules than LPs may also be perceived by plant cells. Medium alkalinization takes part in the plant defence-associated early responses to various biotic elicitors (Felix et al., 1999; Boller, 1995) and we used this readily measurable phenomenon on suspension-cultured tobacco cells to test the activity of products secreted by B. subtilis S499. To this end, supernatant samples were collected at various time-points during growth in the optimized medium. Crude supernatant samples were pre-purified on C-18 cartridge to yield the MeOH40 extract (fractions eluted with 40% methanol) containing molecules with intermediate hydrophobicity, and the MeOH100 extract (fractions eluted with 100% methanol), retaining more apolar compounds among which are found the LPs of the three families. The effect on tobacco cells of C-18 extracts prepared from samples collected every hour all over the culture time were first estimated with pH indicator paper. For clarity purpose, only some of them are represented in Figure 1B.
MeOH40 extract did not induce any change of the pH, suggesting that no metabolite with intermediate hydrophobicity produced by the strain S499 upon such culture conditions are active in triggering an alkalinization response by tobacco cells. The same applies for MEOH100 fractions obtained from samples collected before surfactin production, indicating that apolar metabolic products that could accumulate early in the exponential growth phase are not active on pH. By contrast, pH increases could be clearly visualized by using MeOH100 extracts prepared from culture samples collected then after. A weak but significant pH alkalinization could be observed by treating tobacco cells with extract from 16 h old culture, corresponding with the apparition of surfactin in bacterial supernatant. Stronger pH alkalinizations were observed when MeOH100 extracts from samples containing higher surfactin concentrations were added to the cells. These results were confirmed and more precisely measured with a pH meter (Figure 1C). These first results strongly suggest that surfactin lipopeptides are tightly involved in the induction of this alkalinization process since it
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin fengycins whose production is only 3-4 hours-delayed compared to surfactins (data not shown) could not be ruled out at this stage. By contrast, iturins that accumulate later in the culture are seemingly not involved since iturin-rich samples collected after 72 h of growth did not showed enhanced activity compared to those collected at 54 h with reduced iturin content (data not shown).
Figure 1. Alkalinization response of tobacco cells induced by Bacillus subtilis S499 culture supernatant extracts.
A. Bacterial growth and surfactin lipopeptide production kinetics. Biomass was estimated by optical measurements at 600 nm. Surfactin concentration was determined by HPLC.
B. pH changes, as revealed with indicator paper, in tobacco cell culture following treatment with semi-purified extracts of S499 supernatants collected periodically starting from 13 h to 54 h of incubation.
C. Extracellular pH modification (measured with glass electrode) of tobacco cell culture induced by treatment with S499 supernatant extracts respectively purified after 16 h, 26 h and 54 h of culture.
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
155 In order to get more conclusive information about the relative alkalinization-inducing activities of the three lipopeptide families, we tested MeOH100 extracts from various derivatives of the Bacillus amyloliquefaciens FZB42 strain affected in the biosynthesis of different lipopeptides. The AK3 mutant efficiently produces surfactin but fengycin and iturin biosyntheses are suppressed. The CH1 and CH2 derivatives retain fengycin and iturin synthesis respectively but are impaired in the production of the two other lipopeptide families. LPs production by these strains was qualitatively checked and quantified by HPLC. MeOH extracts were diluted in order to obtain a final LP concentration of 2 µM after addition into the tobacco cell culture medium. A strong alkalinization response of 0.9 pH unit was observed following treatment of tobacco cells with the AK3 extract (Figure 2). Neither CH1 nor CH2 extract induced a significant pH increase compared with cells treated with methanol and used as control. These results confirm that iturins and fengycins are poorly active in triggering an alkalinization response of tobacco cells and that, among Bacillus products, this activity is probably due to surfactins.
Figure 2. Variation in extracellular pH of culture medium of tobacco cells in response to culture extracts from Bacillus amyloliquefaciens FZB42 mutants. The three mutants of strain FZB42 that are affected in surfactin (surf), fengycin (feng) and/or iturin (itu) production are CH2 (surf-/feng-/itu+), CH1 (surf-/feng+/itu-) and AK3 (surf+, feng-, itu-).
Lipopeptides were extracted from culture broths of Bacillus amyloliquefaciens FZB42 mutants by using C-18 solid phase extraction cartridges. Tobacco cells suspension cultures were treated with 21 µL of culture extracts dissolved in methanol 100% to obtain a final lipopeptide concentration of 2 µM. Control consisted of cells treated with same volume of methanol 100%.
Alkalinization as a dose/structure and calcium dependent response to surfactin
On the basis of these informations, we focused on the effect of surfactins and we first used a 99%-pure mixture of homologues produced by the S499 strain to further investigate tobacco cell responses to this elicitor. Surfactin ability to induce a pH shift in tobacco cell medium was first evaluated using increasing amounts. Surfactins added at final concentration of 0.1 µM did not provide any effect on cultured cells but treatment with 2 µM final concentrations triggered
0,00 0,20 0,40 0,60 0,80 1,00
0 2 4 6 8 10
Tim e (m in)
∆ pH
CH1 CH2 AK3 Control
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin markedly in the presence of a surfactin concentration of 4 µM but further increases to 10 µM.
A concentration of 20 µM was less effective than 10 µM suggesting a saturation of the phenomenon as already observed in other studies (Figure 3) (Felix et al., 1999; Bourque et al., 1998). Taken together, these results suggest than even if a low surfactin concentration in the medium can provoke alkalinization, a threshold concentration is seemingly required to induce a maximal cell response.
Figure 3. Dose-response curve for pH alkalinization induced by surfactins on tobacco cell suspensions.
Cultures were treated with purified surfactin from Bacillus subtilis S499 (black curves) to obtain the final concentrations mentioned. In addition to treatment with surfactins (4 µM), tobacco cells were also pre-treated with LaCl3, a calcium channel inhibitor before lipopeptide addition. As purified surfactins were dissolved in methanol 100%, the control consisted in cells treated with same volume of methanol. Representative data from three replicates are presented.
As revealed after separation by HPLC and identification by MALDI-TOF mass spectrometry, the surfactin (all of the A form) mixture from strain S499 is mainly composed of six homologues with C12 to C15 linear saturated acyl chains and two for iso-C14 and iso-C15
forms. These homologues were purified by semi-preparative HPLC and were tested individually on tobacco cells except for the two C13 and the two C15 forms that could not be completely resolved and were tested as single samples. Means calculated from three independent experiments revealed that homologues with the shortest lipid chains (C12 and C13) failed to induce any significant pH shift (Figure 4). By contrast, C14 and C15 surfactins triggered a significant alkalinization response that probably account for most of the activity of the mixture sample. Interestingly, chain ramification may also be important for elicitor activity since a low but significant difference was observed between cells treated with linear (C14n) or ramified (C14i) surfactin homologues (Figure 4B). The synthesis of linear and methylated
Résultats – Chapitre IX – Early responses of tobacco cells to surfactin
157 forms of the C14 surfactin was performed to appreciate the importance of some traits of the peptide part of the molecule. As shown in Figure 4B, both modifications considered individually led to a significant decrease of the eliciting activity on tobacco cells and the linearized and methylated form only retained about 25% of the activity.
Figure 4. Structure-activity relationship of surfactin in eliciting pH alkalinization of tobacco cells medium.
A. Medium alkalinization response to treatment of tobacco cells with surfactin homologues with variable lipid chain from 12 carbons (C12) to 15 carbons (C15). For surfactin with 14 carbons in the lipid chain, a linear chain (C14n) and a ramified chain (C14i) were tested. Cells were treated with purified surfactin dissolved in methanol 100%.
Treatments were added to obtain a final surfactin concentration in cell suspension of 2 µM. The
Treatments were added to obtain a final surfactin concentration in cell suspension of 2 µM. The